Physics
Not being a rocket scientist but Framers idea is quite plausible.
The other alternative as I said earlier is to use toga and those speeds for max TOW - that way ( almost) no take off becomes marginal ( except where you are below max structural- which in Emirates happens only in rare occasions).
Turning a high performance aircraft into a marginal performing one - which is what the current procedures end up doing.
Ask the punters if they want to pay an extra few dollars to use full power - or not... once again like telling them we fly slow and get them to places late when they have paid for a ticket - just to save fuel.
Granted, if you measure all variables with a micrometer - so that you can extend the field length out to a 'balanced field' ( where you can stop or go at V1 within the avail field) then this might save engine life etc but get any one of these wrong ( pax/pallet weights/bit more tailwind at one end of the runway etc) then all of a sudden you have no buffers left.
From basic newtons law stuff - s=ut+1/2 at.t
ut =zero therefore
at any point s=1/2 a. t squared.
a= comes from the accelerometers, therefore if constantly integrated over the changing values of acceleration a figure for time to get to a particular distance is really easy. ( yeah I know that is the trend vector)
All this information is on the jet....position/accel/et etc
I really think it would be possible to use the graphical method that Framer comes up with to come up with a gross error check point.
HOWEVER.... there is already gross error check points ( green dot/vref for the FMC weight etc) and these were ( possibly) not picked up.
Another possible and easy check would be for the performance figures to be pre calc'd for the ETOW by the company and transmitted on the OFP for the predicted take off conditions. or via acars. (Or.... even up linked like some operators)
This should be done on every takeoff to work out regulated anyway ( but it isn't where I work until the pilots do it for most takeoffs)
Would at least give some ball park numbers... but once again it has to be checked by someone ...
I think this is the point of all these possible solutions.... they still require a process to be followed.
Might reduce the swiss cheese holes but still wont stop the problem - just drive the stats out to past 1x 10-9.!
Anyway I really feel for the guys - these mistakes - if that is indeed what happened could happen to anyone of us.
We dont live in a perfect world the more we model it to suit ourselves the more it jumps back at us!